Character & Entity References

Introduction

Any character may be included in an XHTML document using a character reference if its ISO/IEC 10646 code point is known (ISO/IEC 10646 is closely related to Unicode - see *Unicode & ISO/IEC 10646). A code point is usually represented in hexadecimal - for example, the letter 'M' has hex code point 004D - and either this hex value or its decimal equivalent may be used in the character reference. To illustrate: the letter 'M' may be represented by either M (hex) or M (dec). As you can see, a hexadecimal character reference begins with &#x and a decimal character reference begins with only &#. Both are terminated with a semicolon. The x is case-sensitive in XHTML 1.1 documents (it must be in lower-case - see *XML 1.0 Character and Entity References) but the hexadecimal number itself is not, meaning that M and M are equivalent.

Of course, it is much easier for those with a suitable keyboard merely to type the letter M directly. This is not so with more unusual characters. For example, there is a smiley face character (☺) at hex code point 263A which most people will not be able to type directly into a document using their keyboard. Using character references, however, it may be represented as either ☺ or its decimal equivalent ☺. You can look up code points in the *The Unicode Character Code Charts - for instance, the smiley face can be found in the Miscellaneous Symbols chart ("Symbols and Punctuation" section).

It can be hard work looking up or remembering code points for characters. To make life easier, XHTML 1.1 declares three sets of character entities for some more commonly used characters so that a meaningful name may be used to refer to each character within a set (see: Character Entities Defined by XHTML 1.1). For example, the Special Character Entity Set declares an entity (euro) for the Euro sign (€) so that the entity reference € can be used instead of the unmemorable € hexadecimal reference. Note: Entities are case sensitive.

Note also that browser support varies for some of these characters and also depends on the font chosen. Adding 'Arial Unicode MS' to the font-family CSS declaration may be necessary for IE users to view certain characters.

Character Entities Predefined by XML 1.0

Whilst it is usually a matter of choice as to whether to use the character itself directly or a reference to it, some characters must be escaped (using a reference) in order for the XHTML code to be well-formed (see W3C Validation for details of what needs to be escaped). The *XML 1.0 specification (upon which XHTML is based) includes five *Predefined Entities (amp, lt, gt, quot and apos) to allow easy reference to such characters.

You can safely use any of these five entity references (listed in the first column below) in any XHTML 1.1 document, but note that ' is not declared in HTML 4 (you will need to use ' or ' if compatibility is desired). See *Character entity references in HTML 4 for details of which entities are declared in HTML 4.

Character Entities Declared by XHTML 1.1

In addition to the five Character Entities Predefined by XML 1.0, three character entity sets are declared by XHTML 1.1 so that a meaningful name, rather than the decimal (or hexadecimal) character code, may be used to refer to each character within a set. These character entity sets are:

1. Latin 1 Character Entity Set
2. Special Character Entity Set
3. Mathematical, Greek & Symbolic Character Entity Set

<p>Price: &euro;2.25</p>
<p>Price: &#x20AC;2.25</p>
<p>Price: &#8364;2.25</p>

Note: It is only the five Character Entities Predefined by XML 1.0 which are guaranteed to be defined in any XHTML 1.1 document - it is safer to use numeric character references. Entities are case sensitive.

Latin 1 Character Entity Set

Special Character Entity Set

Mathematical, Greek & Symbolic Character Entity Set

Note: FRACTION SLASH might possibly not be rendered if used in isolation (not as part of a fraction).

Other Common Character References

Below are listed numeric references for other common characters. This table is likely to be of limited use, however, since these characters are usually found on a standard English keyboard (and are not used to denote the start of XHTML markup) meaning that they are not commonly escaped.